The testing of water in certified laboratories for both total coliforms (TC) and E. coli at present is usually accomplished using two different tests. Testing of water for drinking and recreation use requires much time. Other samples that are frequently tested for TC and/or E. coli include urine samples (human and veterinary), foods, drugs, and pharmaceuticals. Testing of waste water, aerosols, soil and sludge are sometimes required to evaluate the need for control of harmful organisms.
Drinking water regulations under the Final Coliform Rule require that TC-positive drinking water samples be examined for the presence of E. coli or fecal coliforms. Use of current membrane filter technology to detect total and fecal coliforms necessitates concurrent or serial analyses using two different types of media incubated at two different temperatures. Some of the promulgated prior art E.coli testing methods are confirmatory tests, not primary isolation procedures. The combined procedures (total coliform test and either fecal coliform test or E. coli method) take 28 to 48 hours. The Most Probable Number technology can take up to 72 hours.
Currently, there is no single membrane filter method either in general usage or by approval of the U.S. Environmental Protection Agency that can detect total coliforms and E. coli simultaneously in water.
The media available present several deficiencies. Most, as previously mentioned, detect only one organism or group of organisms, and so require the use of two different tests. For example, they may test for either fecal coliforms or total coliforms or for E. coli. The use of two media analyzed either concurrently or serially will require many resources in time, labor, materials, equipment, and laboratory space. Some methods now used require one specific enzyme substrate to identify one target organism or group and use methods without enzymatic substrate for another group. Hence, two set-ups and types of media are required to meet the requirements of the regulations.
Many of the methods presently employed do not use isolation media. (They do not result in isolation of organisms directly from the sample.) Such tests are used to confirm the identity of organisms isolated on another medium. The over-all result is delay while two-step processes are accomplished to evaluate the extent of contamination.
Several tests use liquid media in a Most Probable Number (MPN) test format that permits the statistical estimation, but not enumeration, of the target organisms. Although the regulations only require the detection of the presence or absence of organisms, enumeration is useful in determining the extent of contamination and in monitoring remediation. The MPN procedure has a built-in positive bias and tends to overestimate the numbers of organisms present. This bias may result in apparent increased compliance violations and rejection of acceptable drinking water.
Many tests use ingredients that are insufficiently effective in recovering the target organisms. Failure of recovery may also result from use of elevated incubation temperatures required by some testing protocols. Elevated temperatures can result in retardation of growth or prevention of the recovery of injured organisms in the sample.
In many instances, the media are useful only for a limited range of samples. For example, it may be necessary to have a different medium for urine specimens than that for water, and a third medium may be needed to test food.
Two commercially available liquid MPN media are available in tests called Colilert and ColiSure. It is stated that both total coliforms and E. coli are detected by these tests simultaneously within 24 to 28 hours. Colilert utilizes 2-nitrophenyl-.beta.-D-galactopyranoside (ONPG) and ColiSure uses chlorophenol red-.beta.-D-galactoside as substrates to test for .beta.-galactoside. Both utilize 4-methyl-umbelliferyl-.beta.-D-glucuronide to test for .beta.-glucuronidase. These media are expensive. The tests may be used to detect presence or absence of target organisms and/or may result in an estimate of numbers of organisms rather than in an enumeration of target organisms. Both tests have been approved by the USEPA to test for total coliforms and for E. coli detection in drinking water. Concern about the high false negative rate of Colilert with disinfected drinking water has been raised by Clark, et. al. (Clark, et. al., Abstract, Annu. Meet. Am. Soc. Microbiol. (1990) Q8, p. 289).
The following definitions are used in relation to substrates for detection of organisms:
A chromogen (or chromogenic substrate) is a substance, (usually colorless) that is acted upon by an enzyme to produce a pigment or dye.
A chromophore is a group on or part of a chromogen that produces a color when the chromogen is cleaved by an enzyme.
A fluorogen (or fluorogenic substrate) is a non-fluorescent material that is acted upon by an enzyme to produce a fluorescent compound.
A fluorophore is a group on or part of a fluorogen that is responsible for the fluorescence when a fluorogen is cleaved by an enzyme. (Galactoside is another term for galacto-pyranoside.)
U.S. Pat. No. 4,923,804 to Ley, et al., teaches use of .beta.-glucuronides to test for E. coli and that the indoxyl-.beta.-D-glucuronide is a preferred agent. (See Ex. 2 of the reference cited). However, he teaches, at column 1, 1. 50-68 that the use of MUG compounds to test for E. coli on a membrane filter test is not appropriate since the fluorescent light can be subject to interference in a membrane filter test. Hence, the teaching of Ley would discourage one from use of an agent having a 4-methylumbelliferyl fluorescent moiety in a membrane filter test. The medium differs from the substrate of the invention in several respects, 1) The medium of Ley can only detect E. coli and does not provide for detection of total coliforms. Because of this, a second medium would be required to identify total coliforms, thereby increasing the time, labor, material, and cost to the laboratory performing the analysis. 2) The base medium of Ley contains glycerol as a nutrient and lacks an inducer and an inhibitor of gram negative bacteria that can give a false positive response. Glycerol in media also causes spreading of colonies making enumeration and discrimination difficult. 3). The medium of Ley is incubated at an elevated temperature (44.5.degree. C.) that would be detrimental to the recovery of injured microorganisms.
U.S. Pat. No. 4,591,554 to Koumura, et al., discloses use of fluorescence analysis using umbelliferone derivatives, including phosphates and galactosides. That reference also teaches use of lactose as an inducer. The organisms are first inoculated into broth for growth. There is no inhibitor in the media, and the reference indicates, at column 3, lines 40-45 that the test also picks up Erwinia, Proteus, and Salmonella--gram negative organisms not usually classified among the coliform bacteria. The media recommended by Koumura can also promote the growth of many other types of organisms such as gram positive bacteria, yeasts and fungi that may also be present in the samples. (See Table 2 of that reference.) Some of the non-coliform organisms are able to inhibit the growth of coliforms. Hence, the tests of Koumura are not appropriate for use wherein there is a desire to find the total coliform populations.
Babelona, et al., (J. Micro. Meth. 12: 235-245) discloses use of MU-glucuronide complexes in testing for E. coli as does much of the prior art. There is no test using a non-fluorescing chromogen-glucuronide to test for E. coli, nor is there any MU-galactoside test for total coliforms.
An agar medium developed by Petzel and Hartman (Appl. Environ. Microbiology. Vol 24: 925-933) used a selective medium for total coliform identification combined with detection of E. coli using 4-methylumbelliferyl-.beta.-D glucuronide. Problems with this medium include the inability to use standard diluents, a high false positive rate when high levels of Flavobacterium species or oxidase positive organisms were present in the water samples, and difficulty distinguishing the natural fluorescence of Pseudomonads from the fluorescence produced during substrate breakdown. Furthermore, this medium could not be used for precise enumeration of the target organisms because of the large number of other Gram negative bacteria that grew on it.
An agar medium using two different enzyme substrates and to a different base medium than those used in the MUGal-IBDG Agar (also known as MI agar) of the invention was developed in Germany to simultaneously detect both total coliforms and E. coli. Total coliform colonies were identified by the production of blue color from the .beta.-galactosidase cleavage of the substrate X-gal (5-bromo-4-chloro-3-indolyl-.beta.-D-galactopyranoside), while E. coli colonies were detected by the fluorescence of 4-methylumbelliferone, produced by the cleavage of 4-methylumbelliferyl-.beta.-D-glucuronide by .beta.-glucuronidase (Manafi and Kneifel, Zbl. Hyg. 189: 225-234). The reference teaches the use of broth or agar containing 4-methylumbelliferyl-.beta.-D-glucuronide (MUG) and a galactoside with a non-fluorescing chromophore (X-Gal). It also teaches a 4-methylumbelliferyl-.beta.-D-galactoside (MUGal) with a glucuronide attached to a non-fluorescing chromophore, 4-nitrophenyl-.beta.-D-glucuronide (PNPG), in an agar medium containing bile salts to inhibit the growth of organisms that are not coliform bacteria. It is reported that the results of use of this media to test drinking water were not good (p. 230). Manafi attributes his difficulties to the color of the drinking water and the reagent. Manafi indicates that effectiveness of reagents on solid and liquid media differs.
Manafi, et al., Microbiological Rev. 55: 335-348 (1991), is a general review article about fluorogenic and chromogenic substrates used in bacterial diagnostics. The particular preferred fluorogenic and chromogenic agents and nutrient substrates used in the invention are not taught therein, and no guidance is provided therein regarding use of inhibitors as required by the instant invention.